On Tue, Apr 30, 2013 at 05:26:50PM +0200, Aram H?v?rneanu wrote:
> > I don't see how they would benefit from being rewritten in go.
>
> Sometimes I need to deploy something written in rc(1) over a
> heterogenous Linux cluster,
This is in part Plan9 related. For kerTeX, simply because my compilation
framework was done this way, I use POSIX.2 utilities: sh(1) and make(1)
mainly. So, on Plan9, kerTeX compiles under APE.
But, despite POSIX being the one standard for systems, not everything is
really present everywhere, even on Unices (ed(1) is typically something
that is left out, while this is the only _line_ editor so the only
editor that should be here).
So I will have, to ensure that kerTeX runs everywhere, to provide some
core utilities. And I think I will switch from sh(1) (ksh or ash) to
rc(1) because rc(1) has regexp manipulations, and this is typically a
lot of what is done in the scripts (the a=$(echo $b | sed ...) is not
really efficient or beautiful).
So there may be a rc(1) base sys utilities set some day. (But if someone
does it before me, I will not be unhappy)
Because, using only C89 (or C99) is great. System(3) is standard C. But
the interpreter is neither guaranteed nor defined. If you have scripts
(and MetaPost has some for example; and kerTeX administration needs some
for the packages...)
--
Thierry Laronde <tlaronde +AT+ polynum +dot+ com>
http://www.kergis.com/
Key fingerprint = 0FF7 E906 FBAF FE95 FD89 250D 52B1 AE95 6006 F40C
